Everything I've Learned About Health, Mapped

I am not a doctor, and this page does not tell you what to do. It is a map of the territory, the six areas that actually decide how you feel and how long you stay functional, with the evidence I trust for each, the small number of levers that carry most of the effect, and what I did about it myself, including where I got it wrong.

It exists because health advice arrives as an undifferentiated firehose in which everything is presented as equally urgent, and almost none of it is. Most of the effect in each domain sits in one or two places. The rest is detail you can safely ignore until the big thing is handled.

Everything below is either a named guideline, a book, or something I measured on myself. Where I am passing on something I have not personally tested, I say so.

Where to start, depending on who you are

Six domains is too many to work on at once, and you do not need to. The evidence below applies at any age. What changes is which part is currently your bottleneck.

Nothing here is age-gated. Sleep, protein, aerobic base and screening matter at twenty-five and at seventy; the emphasis shifts, the fundamentals do not. Where the evidence differs by age, the section says so.

First, the thing that decides whether any of this works

Adherence beats optimality, and it is not close. A mediocre plan you follow for a year outperforms an excellent one you abandon in three weeks, and I have watched this from both sides: as a nutrition coach with forty clients, and more recently as somebody who knew exactly what to do and stopped doing it.

The framing I now use is pull, not push. A behaviour survives when it is pulled by something you actually want: feeling clear-headed in the morning, having energy at 4pm, being able to carry your kid up the stairs. It collapses the moment it becomes a standard you enforce against yourself, because then it is competing with your willpower rather than being carried by your motivation.

The practical version: pick the smallest thing you would still do on your worst week, attach it to something you genuinely want rather than to a rule about discipline, and never negotiate the floor downward. A ten-minute walk that happens every day beats an hour at the gym that happens in your imagination. Book: The Power of Habit, Charles Duhigg. The cue-routine-reward loop is the most useful mental model I have for why a habit sticks or dies.

Every domain below assumes this. If you take one thing from this page, take this one.

Adherence cheat sheet (🔧 all of this is practitioner heuristic, not trial evidence)

• Pick the version you would still do on your worst week. That is your floor.
• Attach it to something you want, not to a rule about discipline.
• Never negotiate the floor downward; never turn it into a ceiling.
• Missed a day? The next repetition is the only thing that matters. Skipping twice is what ends streaks, not skipping once.

Sleep

What the evidence says. Researched · AASM / Sleep Research Society consensus Sleep is the domain with the widest gap between how important it is and how little attention it gets. It is upstream of nearly everything else here: appetite regulation, blood pressure, recovery from training, mood, and how well you make decisions about all of the above. Both DASH-style blood-pressure guidance and general cardiovascular guidance treat inadequate sleep as a genuine risk factor rather than a lifestyle preference.

The levers that matter. Two, in order. Quantity first. Most people who think they have a sleep-quality problem have a sleep-quantity problem, and no amount of optimisation fixes not being in bed long enough. Then timing consistency. A stable wake time anchors everything, and it does more than any supplement or gadget.

What I did. Lived I tracked mine and found the number was the problem: my sleep quality was consistently excellent and my average duration was around five and three-quarter hours, which is a little over half of what I actually needed. No amount of bedroom optimisation was going to close a gap that size. What worked was a cascade of fixed times rather than a bedtime (the kitchen closes at 20:30, screens go off at 22:30, lights out at 23:00) because “go to bed earlier” is a decision you have to make while tired, and a sequence of earlier cues is not.

Sleep cheat sheet (📊 the duration figure is guideline-grade; the rest is practical)

• 📊 Aim for 7 or more hours, per the AASM and Sleep Research Society consensus for adults 18–60.
• Fix the wake time first; it anchors everything else.
• Work backwards with cues, not willpower: last real food ~3h before bed, screens off ~30 min before.
• Cool, dark, quiet. Fix quantity before you spend money on quality.
• Suspect your quality is fine and your quantity is not, until measured otherwise.

Food

What the evidence says. Researched · DASH-Sodium (Sacks, NEJM 2001) · PREDIMED The two dietary patterns with the strongest outcome evidence behind them are DASH, built specifically to lower blood pressure, and the Mediterranean pattern tested in PREDIMED, which has the strongest cardiovascular-outcome trial evidence of any diet. They agree on roughly ninety per cent of the plate: vegetables, legumes, whole grains, nuts, olive oil, fish, and not much processed food. When two independent lines of evidence converge that hard, the overlap is the signal.

The levers that matter. Sodium, and this one is worth attention at any age and whatever your blood pressure currently reads, because it is a slow input rather than an acute one, and the average Indian diet runs well above every guideline. The low-sodium arm of DASH produced the largest drop, and in an Indian kitchen the sodium is not in the salt shaker; it is in the pickle, the papad, the namkeen, the packaged masala and the sauces. Protein, which is where most people are short: sports-nutrition guidance lands around 1.6 g/kg for a training adult, and hitting it from a mostly-vegetarian Indian plate takes deliberate design rather than good intentions. Fibre and volume, which is the part nobody markets, because satiety per calorie is what makes a deficit survivable. The Satiety Index work is the single most practically useful thing I read during my certification.

What I did. Lived I built a plate around those three, with the swaps done at the level of the actual food rather than the food group: lemon and raw onion instead of achaar, roasted makhana and chana instead of sev, spices instead of salt, and a flavour toolkit good enough that nothing tastes like a compromise. That last part is not a nicety. A plate you resent is a plate you stop eating.

Food cheat sheet (📊 sodium and protein are well sourced; the rest is practice)

• Half the plate vegetables, a quarter protein, a quarter whole grains.
• Protein: 1.4–2.0 g/kg/day if you train (ISSN position stand). Most people are well under.
• Cut sodium at the source: pickle, papad, namkeen, packaged masala, sauces. Not the salt shaker.
• Prioritise foods with high satiety per calorie. A deficit you can tolerate is the only one that works.
• Build in the thing you enjoy, on the plan, on purpose.

Book: Is Butter a Carb? by Rosie Saunt and Helen West, two registered dietitians taking apart nutrition myths with the actual evidence. It is the book I would hand to anyone who has been confused by wellness marketing, which is everyone.

Movement

What the evidence says. Researched · Mandsager et al., JAMA Network Open 2018, 122,007 adults Cardiorespiratory fitness (VO2 max, the ceiling on how much oxygen you can use) has one of the strongest associations with all-cause mortality of any measurable thing about you, and moving from the bottom band toward average is where the largest risk reduction sits. Peter Attia’s Outlive organises training into four pillars that I think is the most useful framing available: aerobic base, peak aerobic capacity, strength, and stability. The last one is the one conventional medicine ignores and the one that determines whether you are still moving well at eighty.

The levers that matter. For most people it is simply more easy aerobic volume, because the common failure is training everything at a moderate intensity that is too hard to recover from and too easy to drive adaptation. The mechanism is worth seeing rather than reading:

What I did. Lived I had it measured properly rather than estimated, on a lab test with a mask. The number came back 36.3 mL/kg/min, twenty-sixth percentile for my age, which the report labels “Poor.” I am publishing that because the whole point of measuring is to find out, and because a fit-looking thirty-three-year-old scoring in the bottom third is exactly the finding that would never have surfaced from how I felt. The same test showed my anaerobic threshold sitting unusually high, which is the signature of a training history full of strength and play and completely empty of any structured aerobic block. Small engine, well tuned. So the work now is unglamorous easy volume, and I train inside limits my doctor and my own test results set rather than by feel.

How this shifts with age, because it does. The four pillars matter at every age, but their order of priority moves. In your twenties, aerobic base is the cheapest thing to build and the easiest to neglect. From your forties, strength stops being cosmetic; it is the input to how independent you are later. Past sixty, balance is the one that decides outcomes Researched · Cochrane review · CDC surveillance: falls are the leading cause of injury death in older adults and roughly one in four over-65s falls each year (CDC), and a Cochrane review rates exercise for falls prevention as high-certainty evidence. ⚠ With one correction worth having: in that review, strength training alone did not prevent falls. Only strength combined with balance work did. Nobody is too old to start; the reason to start earlier is that you keep more.

The rule I would give anyone restarting: Reasoned, not tested come back at roughly half of what you think you can do, for the first couple of weeks. I have hurt myself often enough restarting at my old numbers to know that the fitness comes back faster than the tissue tolerance does.

Movement cheat sheet (📊 = evidence · 📖 = framework · 🔧 = my own rule. The sourced table is on the training page)

• 📊 Most of your training should feel too easy. A small slice should be genuinely hard. Avoid the middle.
• Train all four: aerobic base, peak aerobic capacity, strength, stability.
• Get fitness measured, not estimated. It is a stronger mortality signal than most things your annual check-up looks at.
• Restarting after a layoff: come back at about half, for two weeks.
• Grip, hinging and carrying age better than anything you do on a machine.
• 📊 For falls specifically: balance training, not strength alone. A Cochrane review found strength work by itself did not prevent falls.

Breath

What the evidence says. Researched · ⚠ the weakest evidence base in this wing This domain has more enthusiasm than evidence attached to it, so the honest version is narrow, and narrower than I first wrote it.

Nasal breathing has a real physiological basis: the nose filters and humidifies, and the paranasal sinuses produce nitric oxide, a vasodilator, which reaches the lungs during nose breathing. Measured nitric oxide output roughly doubles versus mouth breathing, and small studies show modestly better oxygen tension. ⚠ Small. The oxygenation work involves single-digit numbers of subjects.

Slow paced breathing, around five or six breaths a minute with a longer exhale, is the popular blood-pressure claim, and the evidence is genuinely mixed. One meta-analysis of seventeen studies found reductions of roughly 5.6 mmHg systolic and 3.0 diastolic; a more recent systematic review found no significant effect versus control. ⚠ An earlier version of this page said it “reliably” shifts you parasympathetic. That was too strong and I have corrected it. It is cheap, pleasant and low-risk, which is a fair reason to do it, but it is not established as a blood-pressure treatment, and it is not a substitute for anything your doctor has prescribed.

The levers that matter. Nose over mouth, especially asleep. And a long exhale, because the exhale is the half that does the calming. A four-second inhale with an eight-second exhale, for a few minutes, is the whole technique.

What I did. Lived Two things, one of which was a correction. I use the slow-exhale pattern as an evening off-ramp, which measurably shows up the next morning rather than in the moment. And I stopped holding my breath under heavy load, forcefully exhaling through each hard rep instead, after learning the hard way that breath-holding through maximal effort spikes internal pressure in a way that does not agree with everybody. ⚠ The corollary is a genuine safety point: the forceful, breath-holding practices are not universally safe. If you have any blood-pressure history, that family of techniques is a conversation with your doctor rather than something to pick up from a video.

Breath cheat sheet (⚠ the weakest evidence in this wing. Treat as low-risk experiment, not treatment)

• Nose over mouth, awake and asleep.
• To calm down: make the exhale longer than the inhale. Roughly 4 in, 8 out, for a few minutes.
• ⚠ Slow breathing at 5–6 a minute is the most-studied pattern, but the blood-pressure evidence is mixed: one meta-analysis found ~5.6 mmHg systolic, a newer one found no significant effect.
• Under heavy load, breathe out through the effort. Do not hold your breath through a maximal rep.
• ⚠ Forceful or breath-holding practices are not universally safe. Ask a doctor if you have any blood-pressure history.

Screening

What the evidence says. Researched · Outlive, Peter Attia, a physician’s synthesis, not a trial This is the domain where I would push hardest against the default, and Outlive’s central argument is the clearest statement of why: conventional medicine is superb at rescuing you once disease has declared itself and mediocre at intervening decades earlier, when the trajectory is still cheap to change. Most of what eventually kills people develops silently across twenty or thirty years. The tests that reveal a trajectory generally exist, work, and are not expensive by the standards of what they prevent.

The levers that matter. Get a baseline earlier than anyone suggests, and get the tests that show trajectory rather than only the ones that show current status. In practice that means knowing your actual cardiovascular risk markers rather than a single cholesterol number, knowing your fitness as a measured value rather than a feeling, and having blood pressure measured properly at home over time rather than once a year in a clinic where it is always higher.

What I did. Lived At thirty-three, with no symptoms, I ran a proper battery: a lab exercise test with gas exchange, cardiac imaging, continuous glucose monitoring for a fortnight, and a home blood-pressure log kept daily rather than occasionally. What each one found is between me and my doctors. What I will say is that doing it at thirty-three rather than at fifty was the highest-value health decision I have made, and that nobody suggested it to me. I had to go and ask.

Whatever age you are reading this, the answer is the same: now. In your twenties a baseline is cheap and you will be comparing against it for fifty years. In your thirties and forties it is the point at which trajectories become visible and are still easy to change. Past fifty it stops being optional. The only wrong answer is waiting until something declares itself, because a baseline taken while you are healthy is the thing every future result gets compared against, and you cannot go back and take it later.

Screening cheat sheet (📖 mostly a framework, Attia’s argument, plus my own experience)

• Get a baseline earlier than you have been told to. Your thirties is not too early; it is the point.
• Ask for the tests that show trajectory, not just current status.
• Measure blood pressure properly at home, repeatedly. One clinic reading is close to noise.
• Know your fitness as a number.
• Write down what you were told. A result you cannot find later is a test you did not really take.

Measurement

What the evidence says. Researched Consumer devices are good at a few things and unreliable at others, and the distinction matters more than the brand. Sleep and heart-rate variability tracking from a well-validated wearable is useful for spotting trends; a chest-strap heart-rate monitor is accurate enough to train off; an arm-cuff blood-pressure monitor used properly at home gives a far better picture than a clinic reading. Almost everything presented as a single daily readiness score is a trend indicator dressed up as a verdict.

The levers that matter. Measure a thing because you have a question, and stop when the question is answered. The failure mode here is not inaccuracy, it is accumulating data you never act on and mistaking the accumulation for progress.

What I did, and it is the most useful thing on this page. Lived I believed for roughly fifteen years that I was hypoglycaemic. It explained a familiar cluster of symptoms, I had organised habits around it, and I had never once tested it. So I wore a continuous glucose monitor for a fortnight, ate normally, and logged the symptomatic episodes.

The belief was false. Not borderline: my glucose never once went below the threshold that would define a low, across the entire period, including during the episodes themselves. My variability was in the range you would want. Then I tested the obvious fallback theory, that it was the speed of the drop rather than the level, and that failed too: the sharpest fall in the whole dataset happened on a day I felt completely fine. What actually predicted an episode turned out to be how long I had gone without eating, how badly I had slept, how dehydrated I was, and how much stress I was carrying. Four things, none of them glucose.

Then I took the sensor off and did not replace it. Glucose was a question, the question got answered, and continuing to monitor it would have been a hobby rather than an investigation. That is the whole discipline: instrument, answer, retire. The symptoms were always real. The explanation I had carried for fifteen years simply was not.

Measurement cheat sheet (🔧 method, not evidence. This is how I use devices, not a finding)

• Start with a question, not a device.
• Trends are signal; single daily scores are mostly noise wearing a verdict’s clothes.
• Chest strap for training heart rate. Arm cuff, not wrist, for blood pressure.
• When the question is answered, stop measuring. Collecting data you never act on is a hobby.
• Test the belief you have never tested. That is where the value is.

What to read, and what to trust

Chosen for what will actually educate you on each topic, not for what I happen to have on my shelf. Where something widely recommended has a credibility problem, I say so. A reading list that hides that is worth less than no list.

Start with the guidelines, not the books. They are free, they are written by panels rather than authors with a thesis to sell, and they are the thing the books are interpreting.

On PREDIMED, because the story matters more than the result. The landmark Mediterranean-diet trial was retracted in 2018 because randomisation had gone wrong for around a fifth of participants. The authors re-analysed without the affected participants and republished, and the finding barely moved: roughly a 30% reduction in cardiovascular events. I include this deliberately. A result that survives having its own trial pulled apart is stronger evidence than one that was never questioned, and knowing why you trust something is the actual skill here.

On fitness, and a correction to how I first framed it. In the Mandsager cohort the gap between the least fit and the most fit was associated with roughly an 80% difference in mortality risk, with unfitness comparable to or worse than coronary disease, diabetes or smoking. It is a striking number and I quoted it flatly.

⚠ But it is an association from an observational study, and the best randomised trial is far more modest. The Generation 100 trial (1,567 older adults, five years, randomised, with death as the endpoint) found no significant difference in all-cause mortality between supervised training and a control group following ordinary activity guidelines. The likely reasons (the controls also exercised; 95% of everyone survived; five years is short) are worth understanding, and they are set out on the training plan along with the six filters I now use to rank a study. Exercise is still one of the best things you can do. The honest version of the claim is smaller than the popular one.

Books

• Outlive, Peter Attia. The best available framing of prevention as something you start decades early, and the four training pillars used above. Read it for the framework; it is one physician’s synthesis rather than a guideline, and it is more confident than the evidence in places.
• Is Butter a Carb?, Rosie Saunt and Helen West. Two registered dietitians dismantling nutrition myths with the actual evidence. The best antidote to wellness marketing I know.
• Atomic Habits, James Clear, or The Power of Habit, Charles Duhigg. For the adherence problem, which is the real problem. Clear is more practical; Duhigg explains the mechanism better.

Two popular books to read with your guard up

I would rather flag this than have you find out afterwards.

• Why We Sleep, Matthew Walker. Enormously influential and good at making sleep feel important. It has also been the subject of a detailed, widely-discussed critique documenting overstated and unsupported claims, including a much-repeated cancer statistic and a WHO declaration that does not appear to exist. Read it for the motivation; take the AASM consensus for the numbers.
• Breath, James Nestor. A great read and the reason many people fixed their nasal breathing, including a good deal of what is in the breath section above. But sleep clinicians have pushed back: most of the do-it-yourself techniques have thin peer-reviewed support, and mouth-taping in particular remains genuinely contested. Take the nose-over-mouth principle; treat the protocols as untested.

Everything above, on one page

All seven cheat sheets collected into a single plain-text reference you can keep, print, or paste into whatever you use to organise your life: the one-page health cheat sheet.

It carries the Indian sodium-swap table as well, which is the part I get asked about most and the hardest thing to remember in a supermarket. Nothing to sign up for, nothing tracked, no email required. It is a text file, and it is yours to edit into something that fits you. If you disagree with a line in it, change it; that is a better outcome than following it.

Message me

If one of these six is obviously your weak one, that is the whole purpose of the page. Go and work on that, and ignore the rest until it is handled. If you are not sure which, or you want to talk through the screening question for your own age and situation, message me. I am not a doctor and will not pretend to be one, but I have been through the process of asking for tests nobody offered me, and that is a much better conversation than a paragraph.

Email: abhishek@terrak.ai LinkedIn: Abhishek Chaudhry

What’s changed

• 31 August 2026. Corrected the movement section. I had quoted an 80% mortality difference as though it were settled; it is an observational association, and the best randomised trial on exercise and mortality came back largely null. Both are now given, with the reasons. The full study-by-study reckoning is on the new training plan page.
• 31 August 2026. Corrected the breath section: an earlier version said slow breathing “reliably” shifts you parasympathetic and is used in blood-pressure management. The evidence is mixed (one meta-analysis found a reduction, a newer one found none) and the page now says so. Every cheat sheet is tagged with whether it rests on evidence, a framework, or my own practice.
• 30 August 2026. Every claim now carries a provenance label, and the falls advice was corrected: strength alone does not prevent falls, only strength combined with balance training does.
• 30 August 2026. Added Where to start, depending on who you are, and made the age-dependence explicit in movement, screening and sodium. The first version read as though it were written for someone my age, which was not the intention.
• 30 August 2026. First published.